Certainty and uncertainty in GSG gameplay (in a HoI4 context :) )

[Axe99]

This issue came up a couple of times lately, and thought it might make for an interesting discussion. The main theme it's come up in has been technology research, but the general idea can be applied anywhere - and that is the idea of certainty vs uncertainty when playing a strategy game (and in this case HoI4 ).

The extreme case of certainty is in games like Chess, Go, Draughts and the like, that forego dice or other random elements, and the 'only' uncertainty is the opponent.

At the opposite end, a game like poker or dice is a fairly extreme case of uncertainty, where most of the gameplay depends on random elements that have to be used 'as best they can' by the players.

The uncertainty of opponents is, of course, always the most significant type of uncertainty, but beyond that, other elements of gameplay uncertainty can do anything from introducing interesting but strategically unimportant flavour to providing randomly-generated implausible nerfs.

HoI4, if NFs are considered part of the 'opponent' uncertainty, tends to stick more to 'certainty in gameplay'. We know exactly when a tech will be researched, an NF completed or when a unit will arrive in certain location (perhaps unless the weather changes - I haven't checked this ). In contrast, EU4 has a good deal more uncertainty, with rulers, events and battles all having significantly larger random factors.

Certainty can make for ease of planning, and means that in a game it's more of a case of the player versus the other players/AI, as the player's choices are pitched directly against that of their opponents'.

Uncertainty, on the other hand, can introduce more 'interesting decisions' and flavour to the game (for example, if a medium tank is taking longer than expected to research, a player might have to keep focussing on light tanks, or if a general turns out to be hopeless (or is captured by the enemy or killed in a plane crash), they might need to mix up their command staff). However, the danger of uncertainty is that a player can feel that they lost because they were unlucky, rather than because they made poor decisions.

Which do people prefer? Should HoI4 maintain it's 'heavy' certainty focus or is there room for more uncertainty? Which is more historically plausible? Which is better for MP? Is there a certain or uncertain outcome to this thread? .

 

[Dalwin]

Very interesting topic. I hope it generates good discussion.

For me, the most important element of uncertainty in the entire game is what goes on between Germany and Russia in an historically based MP game. Yes you can be almost completely certain that Germany will attack. Depending on house rules, you may even enjoy some degree of certainty as to when the attack will take place.

What you do not know is the details of how such an attack and defense will be conducted. Both combatants have a lot of leeway in how they construct their forces. I have now been through quite a few of these and no two are quite alike. In our last game, the German gambled on pursuing exclusively light tanks and never got around to fielding mediums or heavies. He also accelerated the timeline by about a year.

The Russian had not seen this coming and was caught completely unprepared for what happened. The irony was that my German ally (I was Japan) offered the tank treaty to the Soviet early, even though he himself already had no intent of taking advantage to rush medium or heavy tanks. The Russian unknowingly leaned into the left hook by declining the treaty. I had to go weeks without making any open comments on that ill fated decision.

I personally would like to see more uncertainty in the research system, both in timing and in the results. I think the game could survive stepping slightly away from leaning on balance based on sameness. There are enough other factors upsetting balance that some variability would add more benefit from variety rather than detriment from randomness IMO.

 

[Meglok]

Not sure how much you can play around with an uncertainty factor in HOI4, mainly because of the short time frame and the coalition structure of the war participants. You can't really have too much uncertainty as to who will be the protagonists or there isn't much of a war. If ai Japan and Italy both decide to join the Allies it is going to be a quick and boring game.

Too much uncertainty in strategy is best left to MP games. Getting the AI to think beyond a choice of go East or West after Poland in Germany's case, or North or South after China in Japan's case, would be tough given the current state of the AI. There just are not too many viable options because failure to do the first basic steps of conquest derails the whole game or leaves a protagonist with their proverbial pants down around their ankles. Japan without the Chinese industry, manpower, and resources is a minor power. Germany without the Polish conquest is usually too weak to fight.

If you leave "historical" on then you know the progression of focus choices every time. Turning the "historical" option off allows there to be some randomness in ai focus choices, and I suspect this will become more interesting and lead to more varied game play as every major's focus options are expanded (as long as the chosen random option combinations don't derail the game). But once the ai chooses to go down a certain line you pretty much know the path and likely results. If all the major and mid-major powers have multiple valid paths then the combinations of choices could be fun.

Two areas where you could perhaps randomize the game is in focus cost and research cost and success. Perhaps if the costs of focus choices were more random and varied so that you didn't know the timing of when the next one of yours or your opponents was going to pop? Do I have 2 days, 2 weeks, or another month before Germany can go around Maginot? As the Germany player I might see that instead of 28 days it will be 60 this time, but the Allies won't know that. Or the fabricate on Yugoslavia focus only takes 2 weeks instead of 70 days this time.

For research, perhaps if you invested 50 or 75% of the required research points you begin to have a small chance of learning the tech? The more you invest the higher your chance goes. And you may not roll successfully sometimes until you have invested 120% of the baseline. It has always bugged me that success in research was automatic and would always work as soon you spent x amount on it. Research was not always guaranteed or successful and neither was the cost. Learning Fighters II a few weeks early and cheaper than normal, or it taking an extra month to figure out how to deploy Marines might throw a little spice into things. I don't think the ai doing random tech checks for all nations at the end of each day would slow the game too significantly, but that might be a consideration. Don't want 2400 to turn into a 15 second pause each day while the ai runs calculations.

Just some off the cuff thoughts. I do like the concept of a little less certainty. I don't want the randomness factor so large that a fumble of a single die roll screws you for the rest of the game.

One item I would like some certainty on, Adding Aluminum in Asia

 

[Gamer_1745]

This is a very interesting question. As I posted before I have been struggling with it. I have coded TRE so then when you complete the Sudeten NF it fires an event chain of events that could (small %) lead to war between Germany and France, UK & Czechoslovakia. Now how many players who either are not ready for a major war will either be upset & restart or just reload to an earlier save? What I want is some uncertainty & drama of the tension at the time & I want the player to make sure they are ready. If the negotiations are going badly for Germany they have the choice of pushing harder (which could lead to war) or backing down.

Would players like this uncertainty? Or does the lack of a peaceful ending to the Sudeten Crisis upset players?

 

[Meglok]

This is a very interesting question. As I posted before I have been struggling with it. I have coded TRE so then when you complete the Sudeten NF it fires an event chain of events that could (small %) lead to war between Germany and France, UK & Czechoslovakia. Now how many players who either are not ready for a major war will either be upset & restart or just reload to an earlier save? What I want is some uncertainty & drama of the tension at the time & I want the player to make sure they are ready. If the negotiations are going badly for Germany they have the choice of pushing harder (which could lead to war) or backing down.

Would players like this uncertainty? Or does the lack of a peaceful ending to the Sudeten Crisis upset players?

I don't mind a small chance of things like the Czechs fight back, because there was a chance that they would have if France has backed them. Having options at major decision points, as long as they are historically plausible options, is imo a good thing that adds spice. Most of the event trees launched by focuses at decision points do have these random factors coded in place already, although I think many of them only come into play if the player is involved.

My Japanese game I am wrapping had the Chinese back down at Marco Polo as an example. All of the sudden I got Beijing and the Qingdao peninsula for free without a war. What I didn't like was there was no ramifications to accepting this and then declaring war on Shanxi and China later. Perhaps the war goals should have disappeared and I should have had to fabricate? Or maybe the United Front formed automatically before I could declare after the land transfer instead of waiting for x days after teh war starts? I guess my point is there should be a little more thought into the ramifications of the "what if?" actually happening.

Add Aluminum in Asia

 

[Gamer_1745]

Good to know @Meglok

And I hope to get others feedback on this.

Another multi-year event chain I am working on is the Fascist & National Socialist groups (real historical groups) in Switzerland and the possibility that given support from Germany & Italy and maybe from Fascist France could break up Switzerland (non-Fascist France has a chance to block it). When a nation supports one of these Fascist groups there is a random chance of 2 different levels of effects. Since this did not happen historically, there was a real level of concern in the Swiss government at the time, the chance it happens is low. Now with the player putting PP into this with a low chance of success would it be upsetting to the player when it fails?


And Add Aluminum in Asia

 

[Dalwin]

Good to know @Meglok

And I hope to get others feedback on this.

Another multi-year event chain I am working on is the Fascist & National Socialist groups (real historical groups) in Switzerland and the possibility that given support from Germany & Italy and maybe from Fascist France could break up Switzerland (non-Fascist France has a chance to block it). When a nation supports one of these Fascist groups there is a random chance of 2 different levels of effects. Since this did not happen historically, there was a real level of concern in the Swiss government at the time, the chance it happens is low. Now with the player putting PP into this with a low chance of success would it be upsetting to the player when it fails?


And Add Aluminum in Asia

I think that depends to a great extent on the likelihood of success. One thing I did not like in earlier versions of HOI was where the espionage system would require one to repeatedly make a not inconsiderable investment that could have been spent on other useful things to get a random chance of 3% or 5% on some effect that quite often was not a major change.

That is an example of too much randomness. It would have been more tolerable if in the case of a failed roll that small percent chance was added to the next one. This would give a cumulative effect that would eventually mean success though the cost and time required might vary greatly.

In the same game system the randomness to the diplomatic act of improving relations was OK. You always got something for your effort but it might be just a little or might be quite a bit more.

 

[Axe99]

Great discussion all . I deliberately left my 'angle' on certainty/uncertainty out of the OP, but my general preference is for historical plausibility, and in the context of HoI4 I'd argue that would call for a bit more uncertainty. However, I'd argue that it's generally important with randomness that there's some player agency involved, at least for the big stuff. Stellaris' tech system does this well - there's a balance between uncertainty and choice. It's also important not to make things to random (one of the issues that's currently the case if you turn off historical NFs in HoI4 - and HoI4's (imo) biggest nod to uncertainty-based gameplay).

For example, I'd definitely be in favour of more uncertainty in the research and focus (timing of completion) systems. As long as the distribution of randomness wasn't too high, then it's likely to 'work out in the wash' over time, but still lead to hopefully interesting decisions to be made along the way. One thing in terms of the research that could add an agency side to it is if there's a possible research delay, the player (or AI) could spend PP to cancel it out (in immersion terms, through more resources at the problem to deal with the issue instead of more time). That way there'd be a compromise between certainty (but potentially at a cost) and randomness (beyond the probability of the event firing in the first place), that drives interesting choices without a significant likelihood of tanking someone's game.

That said, I'd also think it'd be cool if techs had different 'degrees of difficulty'. Something like nuclear might be more likely to be delayed, something like trucks less. Give nukes aren't particularly game-winning combinations at the moment, it woudn't be game-breaking in that example, but it would need to be handled carefully (ie, it wouldn't be fair to deny someone medium or heavy tanks for years on end because of bad luck).

The performance of generals is another area where there's some potential for uncertainty to add to historical plausibility - I'm sure some of the less well-performing generals weren't expected to perform badly before it happened - but I'd say this is starting to get to the situation where it might have a large impact on the game (for example, if someone had a FM in charge of all their Barbarossa forces, and they turned out to be a dud, then could be the difference between winning and losing). I'm also not as sure about the agency 'before the fact' here.

Individual air and land combat results could probably get away with more randomness, but there's so many of them that I doubt it would add to the game, as the randomness would be mostly hidden, and mostly balance out, at least outside of naval combat (which I think needs a decent random element, and already has one).

Another multi-year event chain I am working on is the Fascist & National Socialist groups (real historical groups) in Switzerland and the possibility that given support from Germany & Italy and maybe from Fascist France could break up Switzerland (non-Fascist France has a chance to block it). When a nation supports one of these Fascist groups there is a random chance of 2 different levels of effects. Since this did not happen historically, there was a real level of concern in the Swiss government at the time, the chance it happens is low. Now with the player putting PP into this with a low chance of success would it be upsetting to the player when it fails?

With events, I think there needs to be some kind of 'player expectation management', so that they know where they stand, and have some agency (particularly for players experiencing the events the first time around). So if the player knows it's got a small chance of success, they know the chance they're taking. That's just my take on it though - you should definitely design your mods the way feels best to you .

Oh, and AAIA, because all the cool kids are saying it .

 

[Praetori]

Weather. Extremely important in all aspects of human society and conflict.
A stormy June 44 with no calm weather and D-Day would've been postponed. Record cold and blizzards in '41 and it's quite possible that Moscow would've fallen.
Stormy Pacific year and CVs would've played a less important role etc etc.

 

[Katsuki126]

Random research would be nice. You do not know when the research end, you just have a chance of it ending each month. Then the brand new tank you just researched would have randomized stats based on your doctrines and design company. Should you invest point im a variant, the benefits would be also randomized.
Same goes for generals, you will simply have a notifications of their state of mind (offense, defense, special forces) but they actual level and skills would be randomized and hidden.

I would love such a system, not in vanilla, but a mod that could do this would be awesome.

 

[Dalwin]

@Axe99

I am glad you brought up the performance of generals. This is a thing captured well in some games covering the ACW. The generals all start as unknown quantities. You have to discover through trial and error which ones are better. I think a similar system might work in HOI.

There is one aspect that makes the ACW more suitable in tis regard, politics. The generals there have political affiliation to their home states. Other political functions that involve the states can be affected by how their favorite son is treated. There you also have the concept of promoting generals so that they can have larger commands and of having some leaders earn the right to have such promotions based on battlefield success. These elements are part of such games even if the option for random general skills are turned off.

It can be fun to see a general with a good rating lose favor because he commanded a battle which ended badly.

 

[Gamer_1745]

That's just my take on it though - you should definitely design your mods the way feels best to you .

Yes, I am not going to build something I don't like, but one of the main goals with my mods is education and if people don't like them & don't play them then they don't achieve their goal. My goal is to stay within the realms of plausibility. Some things I think need a level of uncertainty others I think it might be good, but not needed if players will be put off by it. I am really very interested in feedback and not just pushing my views.

 

[Sanny]

The game wouldn't be labelled a Sandbox if it didn't have a degree of uncertainty to it.

 

[safe-keeper]

The law of large numbers is a factor with research for sure. You can gain a -5% penalty here or a +10% bonus there, but if the chances are the same for penalties and bonuses, then it'll always even itself out in the long run. As Axe says, though, of course it might still lead to some interesting scenarios along the way.

I would actually like it if more complex stuff like nuclear bombs or jet fighters were more risky, however, so that you weren't 100% sure at the onset of the war that by '44 or '45 you would have nuclear bombs. You could have the nuclear techs take a very long time to research, and have small and large breakthrough events as well as things like resource availability factor in as well, but in the end, in my eyes it shouldn't be a matter of cetainty, but a gamble you were throwing huge amounts of resources into, in HOI4's case the long-time dedication of one or more research slots and the construction of nuclear reactors. The Allies and Axis both spent much time and energy on developing nuclear weapons, and trying to sabotaging the other side's programs, but only one side actually got the pay-off of nuclear bombs before the end of the war. This would allow nuclear bombs to be made a bit more powerful, too, in my eyes.

Also, I suppose it could be a field that paid off in other ways even if you didn't get the nuke. In vanilla HOI4 you get a research speed bonus when you research the first nuclear tech, but I imagine that you could get other bonuses of various kinds, too. It shouldn't be a black-and-white "either you get the bomb or you get nothing at all" system.

Another cool field to explore would be code-breaking. Right now it seems to be just a numbers game with decryption and encryption techs and maybe a minister or NF or two, I can't remember. So having cryptology as another one of those that might pay off in a big way if you invest heavily in it, along with nuclear bombs, jet fighters, and so on, would be cool.

What I miss the most in HOI and especially HOI4, though, is random events, both the flavour events and the more dynamic ones. These could also explore various alt-history scenarios, such as Churchill dying of a heart attack during the war. Again, there is just so much to explore here, and both events and Vicky2-style inventions could work so well here. Everything from a chance to make a breakthrough in battlefield surgery methods after having fought so and so many battles to requisitioning civilian ships to add to your number of convoys. Having played BICE, I know this can add so much flavour and depth to the game when done correctly.

I agree weather should have a more significant effect, it certainly does in reality.

 

[Serathion]

The biggest uncertainty seems to come from the specialisation of the armies for the time being, which was already refered to in the posts about the German MP situation.
It might be a good, relatively easy addition to the game if the AI would get some sort of variation in its' production profile.

 

[Axe99]

What I miss the most in HOI and especially HOI4, though, is random events, both the flavour events and the more dynamic ones. These could also explore various alt-history scenarios, such as Churchill dying of a heart attack during the war. Again, there is just so much to explore here, and both events and Vicky2-style inventions could work so well here. Everything from a chance to make a breakthrough in battlefield surgery methods after having fought so and so many battles to requisitioning civilian ships to add to your number of convoys. Having played BICE, I know this can add so much flavour and depth to the game when done correctly.

I agree weather should have a more significant effect, it certainly does in reality.

Very much agree on both of these . 'Balanced' random events would make things interesting. The trick is to make them not be 'out of the blue' unless they made sense (for example, Churchill having a heart attack or Yamamoto getting shot down out of the blue makes sense, while requisitioning ships for convoys might be better as an NF, as it's something players should know whether it's possible or not, and be able to plan ahead with).

A bit broader than the topic of this thread, but the degree of 'gameplay uncertainty' across Paradox games is also relatively varied. The likes of CK2 and Stellaris being quite varied (Stellaris has quite uncertain early and late-games, CK2 starts certain but then spirals into higher degrees of uncertainty relatively quickly as uncertain mortality adjusts things), EU4 probably being somewhere in the middle (lots of 'little' uncertainties, but not too many 'large' uncertainties) and HoI4 on the more certain side of the scale (particularly in historical NF mode, but probably also in ahistorical relative to the other three).

Note - I'm not saying this is inappropriate - the shorter timescale of HoI4 would suggest a relatively less uncertain diplomatic progression over the course of the game than a 700-year run of CK2, say .

 

[Meglok]

Very much agree on both of these . 'Balanced' random events would make things interesting. The trick is to make them not be 'out of the blue' unless they made sense (for example, Churchill having a heart attack or Yamamoto getting shot down out of the blue makes sense, while requisitioning ships for convoys might be better as an NF, as it's something players should know whether it's possible or not, and be able to plan ahead with).

A bit broader than the topic of this thread, but the degree of 'gameplay uncertainty' across Paradox games is also relatively varied. The likes of CK2 and Stellaris being quite varied (Stellaris has quite uncertain early and late-games, CK2 starts certain but then spirals into higher degrees of uncertainty relatively quickly as uncertain mortality adjusts things), EU4 probably being somewhere in the middle (lots of 'little' uncertainties, but not too many 'large' uncertainties) and HoI4 on the more certain side of the scale (particularly in historical NF mode, but probably also in ahistorical relative to the other three).

Note - I'm not saying this is inappropriate - the shorter timescale of HoI4 would suggest a relatively less uncertain diplomatic progression over the course of the game than a 700-year run of CK2, say .

Timescale. It is far easier to have "interesting events" in EU4, CK2, and Stellaris because there is more time for the effects to be diluted or addressed. If you start popping events in HOI4 that really effect game play and are beyond the control of players that is a bad thing imo. The timescale of HOI4 doesn't allow you time to address or repair effects from random events.

Edit - Oops, even I can forget to Add Aluminum in Asia

 

[Axe99]

Timescale. It is far easier to have "interesting events" in EU4, CK2, and Stellaris because there is more time for the effects to be diluted or addressed. If you start popping events in HOI4 that really effect game play and are beyond the control of players that is a bad thing imo. The timescale of HOI4 doesn't allow you time to address or repair effects from random events.

It depends on the scope though. Generals killed in an air accident should be recoverable, while a 20 per cent penalty to IC might not be. It should be feasible to scale the impact to make it relevant to the scale of the game .

More importantly though, didn't you miss something in your post? *cough* AAIA *cough* .

 

[Meglok]

It depends on the scope though. Generals killed in an air accident should be recoverable, while a 20 per cent penalty to IC might not be. It should be feasible to scale the impact to make it relevant to the scale of the game .

More importantly though, didn't you miss something in your post? *cough* AAIA *cough* .

Even that example could be devastating to an Italy or Spain, if it is your only decent Field Marshall you have nursed up thru 1940 to level 6 or 7, only to lose him to a random event. You just lost 25% combat bonus, plus probably his offensive, defensive, and log bonuses. For smaller countries that would be a heavy penalty to pay for random flavor. This is probabl;y why PDS stuck with news events with no ramification at launch.

One thing that shouldn't be random, and that is Add Aluminum in Asia.

 

[Axe99]

Even that example could be devastating to an Italy or Spain, if it is your only decent Field Marshall you have nursed up thru 1940 to level 6 or 7, only to lose him to a random event. You just lost 25% combat bonus, plus probably his offensive, defensive, and log bonuses. For smaller countries that would be a heavy penalty to pay for random flavor. This is probabl;y why PDS stuck with news events with no ramification at launch.

One thing that shouldn't be random, and that is Add Aluminum in Asia.

While it would be a significant factor for a small nation, I'd still argue that it would be overshadowed far greater amounts of uncertainty (the behaviour of all the larger nations). If that 25% bonus for one leader is the difference between victory and defeat, there will be far greater uncertain factors that are more likely to push things one way or the other. It also wouldn't be in isolation - having an event like that might also be a good reason to also have an event like "leadership talent unearthed" in the mix, that adds a competent general, for example.

That said, I'd be all for having these optional, noting that every extra 'option' makes balancing by the devs that much harder.